The proposed study will investigate peer influence on adolescent sexual risk, dating aggression, and substance use by directly examining talk between peers aged 16 to 19. Adolescence is a period of increased involvement in romantic and sexual relationships, which are salient experiences in the transition to adulthood, but also introduce vulnerabilities. In particular, unintended pregnancy, sexually transmitted infections (STIs), and dating aggression are prevalent problems in adolescence that pose dangers to long-term physical and mental health and incur considerable burden to public resources. Other risk behaviors, especially substance use, appear to share common developmental pathways. Often overlooked in the adolescent risk literature has been the transmission of information and influence between peers on these issues, given that friendships also assume more importance in adolescence. Preventive programs that target adolescent risk behavior have increasingly focused on changing peer group-based norms and interpersonal behaviors, but more basic science is necessary to understand the mechanisms and effects of peer influence so that they may be harnessed for improved intervention. The proposed study builds upon an ongoing longitudinal project that provides multiple waves of data on adolescents' risk behavior as well as exposure to family-of-origin aggression, treated here as a predictor of later deleterious outcomes. An ethnically diverse, community sample of same-sex friends will participate in videotaped, ecologically valid discussions of topics important to their lives, including dating and substance use. These candid discussions allow for measurement of cognitive, affective, and behavioral processes of peer influence. Processes will be measured through coding by human raters and also innovative computerized speech signal processing to measure emotional tone from acoustic characteristics of the voice. Longitudinal and multilevel analyses with 131 peer dyads will be used to test mediating and interacting effects, with aims to (1) examine associations between peer talk and risk, specifically dating aggression, sexual risk, STIs, pregnancy, and substance use; (2) test mediating and moderating roles for peer talk in the link between family-of-origin aggression and later risk and romantic relationship quality; and (3) explore gender differences in the effects of peer talk on outcomes. The proposed study will bridge literatures on: the intergenerational transmission of aggression; adolescent dating aggression, sexual risk behavior, and substance use; observational coding and computerized analysis of dyadic interaction; and adolescent health promotion. We argue that greater understanding of peer talk influence is critical to the design of more effective interventions for reducing teen pregnancy, STI transmission, dating aggression, and substance use among adolescents.